Submitted as: model evaluation paper 15 Mar 2021

Submitted as: model evaluation paper | 15 Mar 2021

Review status: this preprint is currently under review for the journal GMD.

Convolutional conditional neural processes for local climate downscaling

Anna Vaughan1, Will Tebbutt1, J. Scott Hosking2,3, and Richard E. Turner1 Anna Vaughan et al.
  • 1University of Cambridge, Cambridge, UK
  • 2British Antarctic Survey, Cambridge, UK
  • 3The Alan Turing Institute, UK

Abstract. A new model is presented for multisite statistical downscaling of temperature and precipitation using convolutional conditional neural processes (convCNPs). ConvCNPs are a recently developed class of models that allow deep learning techniques to be applied to off-the-grid spatio-temporal data. This model has a substantial advantage over existing downscaling methods in that the trained model can be used to generate multisite predictions at an arbitrary set of locations, regardless of the availability of training data. The convCNP model is shown to outperform an ensemble of existing downscaling techniques over Europe for both temperature and precipitation taken from the VALUE intercomparison project. The model also outperforms an approach that uses Gaussian processes to interpolate single-site downscaling models at unseen locations. Importantly, substantial improvement is seen in the representation of extreme precipitation events. These results indicate that the convCNP is a robust downscaling model suitable for generating localised projections for use in climate impact studies, and motivates further research into applications of deep learning techniques in statistical downscaling.

Anna Vaughan et al.

Status: final response (author comments only)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on gmd-2020-420', Anonymous Referee #1, 06 May 2021
  • RC2: 'Comment on gmd-2020-420', Anonymous Referee #2, 25 May 2021

Anna Vaughan et al.

Anna Vaughan et al.


Total article views: 387 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
261 119 7 387 1 3
  • HTML: 261
  • PDF: 119
  • XML: 7
  • Total: 387
  • BibTeX: 1
  • EndNote: 3
Views and downloads (calculated since 15 Mar 2021)
Cumulative views and downloads (calculated since 15 Mar 2021)

Viewed (geographical distribution)

Total article views: 334 (including HTML, PDF, and XML) Thereof 334 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
Latest update: 22 Jul 2021